Advancing characterization of spatiotemporal dynamics of dissolved organic matter after landscape disturbance to inform and protect drinking water treatability

Abstract

Surface waters are critical sources of drinking water globally. The quality and availability of water from these sources is threatened by landscape disturbances (e.g., urbanization, agriculture, resource extraction) including climate shocks (e.g., wildfires, intense precipitation). To ensure reliable and consistent supply of drinking water, offline reservoirs are often used for raw water storage prior to treatment. While largely designed for managing water quantity and availability, reservoir storage can also affect water quality. In addition to providing bypass capacity during contamination events (e.g., accidental spills/releases), hydrodynamic factors such as dilution, physico chemical and biological transformations occurring within reservoirs, including photodegradation, flocculation, and microbial primary production, can alter water quality relative to inflowing source water. Drinking water treatment needs and performance are driven by many aspects of source water quality, one of the most critical being organic matter. Landscape disturbances can substantially change organic matter concentration and character in source water because they alter hydrologic connectivity, mobilizing organic matter from the landscape and affecting its delivery to and transport within receiving waters. Rapid shifts in source water quality can challenge water treatment plants (WTPs), potentially increasing operational costs or leading to service outages. Raw water reservoirs can help attenuate such source water quality change; however, this is not typically done by design. Here, ultraviolet absorbance at 254 nm (UV254)—a real-time indicator of dissolved organic matter (DOM) concentration and aromaticity—was intensively monitored in a wildfire-impacted drinking water system with two off-line reservoirs in series to demonstrate how engineered storage can be leveraged to dampen landscape disturbance-associated changes in source water quality and increase treatment resilience. High-frequency sampling of UV254 did not show meaningful differences across depth, lateral position within the reservoir cells, and time of day. Notably, UV254 standard deviation was 1.6 and 2.2 times lower in the first and second reservoirs, respectively, than in the river source; respective peak UV254 values were 1.45 and 1.6 times lower, indicating effective attenuation of DOM from the river by the reservoirs in series. This attenuation also reduced polyaluminum chloride coagulant dosing by approximately 70 mg/L during a period of particularly deteriorated source water in 2023, helping reduce aluminum residuals and the potential for exceeding regulatory maxima in treated water. Accordingly, this work demonstrates that raw water reservoirs can be intentionally and strategically designed to attenuate disturbance-driven fluctuations in source water quality, thereby enhancing drinking water system resilience and supporting climate change adaptation. Watershed-scale source water quality monitoring is essential for assessing cumulative watershed effects to enable effective management of drinking water supplies and detection of threats to water treatability. However, it can pose substantial logistical challenges. Source water quality, including the concentration and character of organic matter, can change during the unavoidable period between sample collection and analysis, making effective sample preservation critical for maximizing return on monitoring effort. While a standard method is available for the preservation of dissolved organic carbon (DOC), a common surrogate for DOM, its implementation is not always feasible. Difficult access to remote or hydrologically critical headwater locations for sample collection, limitations on in situ sample preservation, and associated challenges in transporting collected samples to analytical facilities within prescribed holding times can compromise data integrity and representativeness. Several alternative preservation (e.g., filtration, freezing) techniques for organic carbon concentration and character are often used when adherence to Standard Methods is impractical; however, their efficacy is highly variable and has not been systematically evaluated. Thus, the direct and combined effects of temperature (e.g., refrigeration, freezing), filtration, and acidification were systematically investigated here using four natural water matrices of diverse quality. This analysis revealed that freezing effectively preserved DOC except when combined with acidification (which would be suggested by extension of Standard Methods); the combination of these methods caused significant shifts in DOC in some cases. Specific ultraviolet absorbance at 254 nm (SUVA) changed within 24 hours of sample collection in some water matrices; this was also reflected by change in the humic substances fraction of DOM, which was measured by size-exclusion chromatography using liquid chromatography-organic carbon detection (LC-OCD). DOC in unpreserved water samples remained stable (i.e., < 10% change) for up to seven days after collection. Accordingly, while Standard Methods remain the benchmark for preserving DOM in water samples, this work advances practical strategies to maximize the value and interpretability of monitoring data at conditions where adherence to these methods is not feasible. Collectively, these findings highlight the value of offline raw water reservoirs and robust sample preservation strategies for enhancing WTP resilience to source water quality change. Reservoirs can effectively attenuate rapid shifts in DOM concentration and character following climate shocks such as wildfires and substantially reduce extremes in coagulant demand. Understanding the stability of DOM during the unavoidable period between sample collection and analysis further helps maximize the value of raw water monitoring to inform water quality and treatability change without compromising data integrity. By integrating high-frequency reservoir monitoring with matrix-specific preservation practices, drinking water utilities can better anticipate source water variability, prepare for disturbance, and improve overall treatment performance and resilience.